Embodied Intelligence for Soft Robots: Proprioception, Perception, Learning, and Control with Advanced Sensing and Actuation

Recent advances in robotics are increasingly driven by the integration of intelligent algorithms, advanced sensing technologies, and novel actuation mechanisms. In particular, the emergence of soft robotics has introduced new opportunities for safe and adaptive interaction with complex environments, enabled by compliant materials and innovative actuation methods e.g., pneumatic actuators. At the same time, progress soft sensor technologies such as optoelectronic sensors has enhanced proprioception capabilities, while machine learning and model predictive control have improved decision-making and system performance. Despite these developments, unifying perception, perception, learning, and control across both rigid and soft robotic systems remains a significant challenge, especially under uncertainty and in dynamic real-world conditions.This Research Topic aims to advance the development of intelligent robotic systems by integrating modern sensing technologies, learning-based methods, and control strategies with emerging soft robotic platforms. The goal is to explore how robots both rigid and soft can achieve robust proprioception, perception, adaptive behavior, and efficient control in uncertain and dynamic environments.We seek contributions that investigate the synergy between machine learning techniques and model-based approaches such as model predictive control, particularly in systems involving complex actuation and sensing. Special emphasis will be placed on soft actuators, vacuum-based actuation mechanisms, and their control challenges, as well as the role of optical and multi-modal sensors in enabling accurate environment understanding.Additionally, this Research Topic encourages work on bridging theoretical developments with real-world applications, including autonomous systems, human-robot interaction, and adaptive manipulation. By combining insights from soft robotics, sensing, and intelligent control, this topic aims to push forward the next generation of versatile and resilient robotic systems.This Research Topic welcomes original research articles, reviews, and perspectives addressing the integration of sensing, learning, actuation, and control in modern robotic systems. Topics of interest include (but are not limited to):• Soft robotics and soft actuator design• Vacuum-based and compliant actuation systems• Optical sensors and advanced sensing technologies• Sensor-actuator integration in robotic systems• Machine learning for perception and control• Model predictive control in robotics• Multi-modal perception and sensor fusion• Intelligent control of soft and rigid robots• Autonomous systems in dynamic environmentsWe particularly encourage submissions that demonstrate experimental validation, real-world deployment, or interdisciplinary approaches combining soft robotics, AI, and control. Contributions addressing both theoretical and applied aspects are welcome. Topic editors: Nabih Pico, Babar Jamil, Altair Coutinho, Vicente Parra-Vega Source: https://www.frontiersin.org/research-topics/79667/embodied-intelligence-for-soft-robots-proprioception-perception-learning-and-control-with-advanced-sensing-and-actuation